Geostatistical Modelling and Web‐Based Mapping of Malaria Risks Among Children Under Five Years in Nigeria: Evidence From the 2021 Nigeria Malaria Indicator Survey
Justice Moses K. Aheto, Bakare Emmanuel Afolabi, Dolapo O. Oniyelu, Deborah O. Daniel, Steven I. Ikediashi, Oluwaseun A. Mogbojuri, Ronke D. Olorunfemi, Sodiq A. Orogun, Afeez Abidemi, Idowu I. Olasupo, Samuel A. Osikoya, Samuel A. Adeyemi, Hapiness O. Ismail, Isaac Bakare, Samson O. Olagbami, Dolapo A. Bakare, Baiyeri Samuel, Temitayo V. Irewole, Oluwafunmilayo O. Olapade, Oluwafolakemi Odunola, Oghenekevwe R. Ajewole, Joshua P. Ojo, Segun Oyedeji, Wisdom TakramahABSTRACT
Background
Malaria is a critical public health concern in Nigeria with the country bearing an uneven burden of the disease. In Nigeria, malaria is one of the main causes of child mortality and despite all efforts to reduce malaria mortality rates, the disease remains a major concern, notably among children under‐fives. There is a paucity of data on more localized predictive malaria risk geospatial maps to inform control and elimination strategies amidst limited public health resources in this setting. This modelling study therefore sought to understand, predict and map malaria risk in the presence of environmental factors in Nigeria.
Methods
This study utilized data from the 2021 Nigeria Malaria Indicator Survey (NMIS), conducted under the Demographic and Health Surveys (DHS) Program. The 2021 NMIS marks the third malaria indicator survey carried out in Nigeria, following previous surveys in 2010 and 2015. The outcome variable of interest is the number of individuals in each sampled cluster who tested positive on the rapid diagnostic test (RDT). This study investigated spatial risk factors for malaria prevalence in Nigeria through geostatistical modelling approaches. The implementation of the models was carried out with the integrated nested Laplace approximation (INLA) method via the stochastic partial differential equation (SPDE) approach in R‐INLA.
Results
The study identified aridity (log‐odds = −0.0400, 95% CrI = −0.0610, −0.0190) and enhanced vegetation index (log‐odds = 9.3930, 95% CrI = 7.4220, 11.3760) as significant predictors of under‐five malaria risk. The fitted Bayesian geospatial spatial model with covariates predicted malaria prevalence of 24.9% with a range of 0.5% to 74.2%. The predicted malaria prevalence was highest in parts of Zamfara (prevalence > 70%) and Kebbi (prevalence > 60%) States.
Conclusion
These findings are of the utmost significance for policymakers involved in malaria control and elimination efforts. They provide evidence‐based information that can guide resource mobilization and targeting, intervention design, and monitoring strategies. The identification of environmental predictors like aridity and enhanced vegetation index suggests the need for location‐specific interventions according to environmental determinants of malaria transmission.